Psychology And Economics: Theory

Starting with the phrase "Economic analysis is based...", the material below dates largely from around 2004. My approach to research in this area has changed considerably in the meantime, partly in response to the debate previewed below in relation to revealed preference having finally burst into the open. Faruk Gul and Wolfgang Pesendorfer wrote a very provocative defense of "Mindless Economics" that was of particular interest in light of the fact that for the last three years I have been running a joint seminar with the distinguished NYU neuroscientist Paul Glimcher. Together with Mark Dean of NYU, I have been working on an appropriately disciplined methodology for introducing rich new data into economic theory. Somewhat against my earlier instincts, we have become really hard-nosed about this, using an explicitly axiomatic approach to modeling novel "psychological" data. We have applied this tough discipline even to neurological data, thereby developing a new approach to neuroeconomic theory. Here is the first formal article in the area, which is forthcoming in the Quarterly Journal of Economics. Mark and I are currently working with Glimcher and his student Robb Rutledge in implementing the called for testing protocols.

Dopamine, Reward Prediction Error,and Economics

The following article outlines the broader methodological proposal on axiomatic modeling of new psychological data. This is an area in which research is ongoing, and which I find intellectually compelling. I will post additional papers as available.

Economic Theory and Psychological Data: Bridging the Divide

The above article is one of 15 in a forthcoming OUP book that Andy Schotter and I are editing,"METHODOLOGIES OF MODERN ECONOMICS VOLUME 1: FOUNDATIONS OF POSITIVE AND NORMATIVE ECONOMICS". This volume will open with the article by Gul and Pesendorfer, a fact for which we have the Journal of Economic Literature to thank. Andy and I are convinced that methodology is a subject whose time has come, and we are going to edit several more such volumes to drive the point home. We are lucky to be working at the Center for Experimental Social Science at NYU, an institution that tries hard to nourish our sense of intellectual adventure.

Another major change in research path involves my taking an increasingly formal approach to survey responses. In large part driven by the same desire to retain strong connections with the analytic core of the discipline, I have been working on survey designs that connect directly with dynamic economic theory. The corresponding "Strategic Survey" methodology is outlined in the following paper, in which it is applied to a question of fundamental importance concerning patterns of spending and saving in the retirement period. This work is joint with John Ameriks of Vanguard, Steven Laufer of NYU, and Stijn van Nieuwerburgh of NYU Stern School. Being novel, it is provoking the usual defensive response among various incumbents. Yet on the positive side, the distinguished Michigan economists Miles Kimball, Matt Shapiro, and Bob Willis, all of whom have been pursuing similar methodologies, have been more than welcoming. They placed me on the Steering Committee of the HRS in part as a result of my work in this area, which may enable me to influence methodology in this area. In fact, survey methodology will be featured in the third volume of the OUP Methodology Series that Andy and I are editing.

The Joy of Giving or Assisted Living? Using Strategic Surveys to Separate Bequest and Precautionary Motives

I anticipate many exciting developments in this body of survey work over the next few years. In terms of method, I will look to tie survey responses ever closer to economic theory. In terms of substance, the retirement area is full of challenges relating to product design, cognition, and pure human drama. As tragic as is the current sub-prime mortgage crisis, it is as nothing compared with the tragedies that will unfold in the area of long term care. Unless something changes, millions of elderly Americans will soon find themselves with inadequate resources for appropriate long term care. Judging by past precedent, the political class will have failed to put in place the appropriate preventive structures, and private businesses will have been unable to find adequate solutions, in part due to obscure regulatory constraints. We will then be subjected to a long parade of self righteousness by those whose present inactivity will be centrally to blame. What a depressing prospect. Fearing another such unbearable display, I am working with a group of like-minded individuals to bring attention to the looming problems and to develop new solutions at this early juncture.

Economic analysis is based in large part on the assumption of deliberative goal pursuit. The choices we make are seen as rational, in light of our goals and our perceived constraints. This assumption has proven extraordinarily fruitful in classical economics, yet is seen by many as meeting its demise in the field of behavioral economics. Behavioral economics aims to introduce certain aspects of psychological realism to decision making. In general, this realism is seen as contradicting the assumption of rational goal pursuit. In research with John Leahy, I have questioned whether the dividing line between economics and psychology should be framed in terms of rational versus irrational. We believe that psychological forces such as fear of the future, suspense, and disappointment, can rationally be factored into many decisions. To accommodate this insight, all that is needed is a recognition that prizes can have a psychological as much as a physical nature. We developed a simple theory to capture this idea. The attached links are both to the article as published, and to the prior working paper, which contains additional illustrations, including an application to gambling.

Psychological Expected Utility Theory and Anticipatory Feelings, with John Leahy, Quarterly Journal of Economics, 2001, 55-80.
Earlier unpublished version

Why would such an obvious idea require elaboration and defense? Because it contradicts a fundamental axiom of the profession: the principle of revealed preference. According to this principle, the only legitimate basis for a theory of choice is direct observation of choices. Mental states are presumed to be unobservable and therefore unsuitable building blocks for the theory of choice. While ordinary folk may discuss fear and anxiety, the principle of revealed preference says that social scientists must limit ourselves to analyzing behavioral consequences. In standard economics, actions speak louder than emotions. In order to widen support for our view that it is time to take a closer look at mental states, it is necessary first to analyze the limitations of the revealed preference view of the world. The following paper makes a start on just this program:

The Supply of Information by a Concerned Expert

The paper shows that there are crucial policy questions that can be answered only if one uses a utility function that goes beyond revealed preference. These are questions involving the supply of information by a well-informed policy maker. The essential idea is this. If dislike of information is the result of an individual's love of surprise (as when that individual is watching a game played by a favorite team without knowing the result ahead of time), it is never a good idea to reveal the actual outcome, even if you happen to know it, and the outcome accords with their preferences. On the other hand, if their dislike of information results from a fear of getting bad news about, say, their health status, then it may be profoundly beneficial to pass on good news as soon as possible. Apparently, the motive for information avoidance is critical in this decision, not merely the fact of such avoidance.

We wrote a second paper that makes an important point about the limitations of the principle of revealed preference. The classical approach involves using choices as a proxy for welfare, and therefore deriving from our models strong policy statements. It turns out that the argument for this approach is far weaker in dynamic than in static settings. The reason is simple: it is not feasible to change past choices, even were such changes to be currently desirable. One can disagree with one's prior choices, yet find no market in which to express this disagreement. Recognition of this problem has a devastating impact on the standard arguments of dynamic welfare theory.

The Social Discount Rate (with John Leahy), NBER Working Paper 7983, 2000.

Armed with these insights, John and I have become increasingly comfortable that choice theorists will ultimately move in our direction. Just as game theorists discuss strategies in a manner based on intuitive realism rather than actual observation, so one day will choice theorists. The following papers move this agenda forward. All three papers focus on one of the profound advantages of our fully rational approach to psychology over partly rational behavioral models. Our approach allows one to conduct policy analysis in areas that have classically been considered too "fuzzy" for such analyses to be worthwhile. In essence, we are able to propose new policies in settings in which factors such as anxiety and fear play a role. We believe that these ideas may be applied fruitfully in the fields of health care, and in the arena of savings, investment, and portfolio choice. Hints (and more) in this direction can be found in the following three papers:

Aids Policy and Psychology: A Mechanism Design Approach, with Kfir Eliaz, forthcoming Rand Journal of Economics

Behavioral Policy(with John Leahy), in Essays in Economics and Psychology (I. Brocas and J. Carrillo eds.), Oxford University Press, 2003

Fear as a Policy Instrument, in Time and Decision (George Loewenstein, Daniel Read, and Roy Baumeister eds.), Russell Sage, 2003

One big question concerning this line of research concerns evidentiary issues. If one is going to go beyond choice, where is the additional information to come from? It is with this question in mind that John and I have pursued survey research, in combination with John Ameriks of the TIAA-CREF Institute, who had his own earlier struggles against the tyranny of revealed preference. According to many economists, the reason that some individuals save far more than others with similar economic and demographic characteristics is to be found in the fact that they have different "discount factors". When John enquired as to the meaning of this elusive concept, he found it to be circularly related to whatever it was that made one person save more than another. Our line of joint research is detailed on the survey research portion of this Web site.